The Mathematical Danger of Democratic Voting

[@auedpo]

6:34 "By playing voters against each other, we, the agenda setters, have caused them to agree on a policy that is much worse for everyone than the policy they began with." Powerful stuff.

[@dnfluffles772]

@@kitkat47chrysalis95 biden, I'm sure there have been more examples throughout history😅

[@Lawrencelot89]

@@kitkat47chrysalis95 Using the term woke is a perfect example of 'playing voters against each other', making it seem it's a battle of woke vs. anti-woke, while in reality the battle should be the powerful vs. the common people.

[@SachinJavsen]

@@kitkat47chrysalis95 Then you must follow Indian Political History to know its true meaning. We have been fooled for like 40-50 years in the name of Gandhi by one political party. Indians believed that political party is doing good but post 1990, Indians come to know how they got fooled and deprived of betterment they would have got. Now, that political party is struggling to make ends meet on voting booth.

[@gdiwolverinemale2745]

@@SachinJavsen lol. I had a debate with a stupid from the subcontinent. He was praising Indira. What a moron

[@maaren3150]

​@@kitkat47chrysalis95 define "woke"

[@theprodigyfmwm7509]

A mathematical model for the race to the bottom of the barrel. This is incredible! Applying such a model to, say, trade or tax policy may be extremely revealing. Thanks for making this video.

[@nomansbrand4417]

Interestingly, this method would not work, if you have lots of voters with continuously distributed opinions about the optimal agenda. Because then a shift from the "average" would never be in favor of the majority. You need to have people being separated in groups, similar to single people above. Another reason to not think in extremes, but to allow intermediate positions!

[@Kevin-cf9nl]

@@nomansbrand4417 Unfortunately, for most policies, continously distributed positions don't work, because incoherent policies don't work. People who favour functioning policies to non-functioning policies will by necessity cluster into islands - that's not the result of extreme thinking, just of understanding the policies are only desirable if you think they'll work. Consider the classic simple policy of where to eat for dinner among a group of friends. There are compromise policies where perhaps the group all orders a bit of food from multiple different restaurants, but that greatly increases the delivery costs - no one wants that. There's also compromise positions like not getting any food at all that are equally undesirable.

[@user-xl5kd6il6c]

@@Stevie-J that's unrelated to everything posted above also, in such a pool of possible things to go against, you chose Ads? topkek

[@cjjuszczak]

@@Stevie-J "Ad buyers are the customer. The customer (pharma) is always right," Wait, you just said there were TWO customers lol

[@lynth]

"Democracy" isn't the problem. It's just that Western capitalist "liberal democratic" systems aren't set up to be democratic but to promote the interests of the capital-owning elites while legitimizing them through electoralism. The existence of party politics and the legality of lobbying and capitalist ownership of media, etc. are all inherently anti-democratic. 1. You can't have democracy under capitalism => As long as capital holds independent power in society, those who own capital (i.e. property generating a passive income enabling its owners to not having to work) will necessarily have a disproportionate ability to influence policy-making. 2. You can't have democracy as long as "opposition parties" exist => If politicians don't all sit at the same table in the same "party" (i.e. the national state) with the same voice, then politics will become a contest for competing special interests. You either need one party committed to democratic leadership (i.e. a socialist vanguard party) or you must abolish all parties outright. 3. You can't have democracy as long as lobbying is legal => If the haves can influence lawmakers more than a poor person, you will not see democratic decision-making. Money being protected as a type of speech is a complete death knell to democracy. 4. You can't have democracy as long as media, schools, universities or any other type of opinion-shaping institution can be privately owned => This one should be the most obvious, but if a bunch of rich people can determine what people can hear in media and get taught at schools, then people will be influenced by that propaganda which will naturally go against their own self-interest. Media and education must be overseen by a democratic control apparatus. 5. You can't have democracy as long as the economy isn't owned by the workers themselves and run democratically => If workers can't dictate their own labour rights and don't own their own companies, then capitalist owners will able to control them and have disproportionate control over company decision-making. All workers need to be part of unions, no company must be influenced by non-worker shareholder interest. All company decisions must be made by the worker collective. Capitalists, obviously, will tell you the exact opposite of all of these things is the case and that they represent freedom, democracy, and self-determination. That is a lie. They don't. Capitalism is antithetical to democracy, freedom, and self-determination. It's a system designed to maintain the power of rich elites and keep the working class in perpetual oppression.

[@Chris-hr8lt]

"Voters don't necessarily blindly vote for whatever policy is closest to their views" Damn I wish you were right xD

[@Redpoppy80]

Yeah... That is not a good assumption to make. There are many factors that can lead to voting for "lesser evils" as this is called politically. Just off of my head, lack of political education, desperation, fear of something worse, being propagandized to name a few. But voting for a lesser evil is only a beneficial thing if the political actors are honest about the agendas being pushed, which they are almost never honest about so how do you want to be screwed?

[@justcommenting5117]

I think this part was ironic considering the way he ends the video

[@alejmc]

Exactly. People in my experience DO vote with whatever is closer to them: “somebody else not me pay for this stuff with other (not mine) people’s money!” time after time after time. And also *crabs in a bucket* style: people prefer to suffer and be in misery as long as that means that there won’t be outliers overly benefiting even if the overall outcome would have been better in that situation. Case in point, where I live, there’s a lot of “let’s tax the rich this and that and more”, the usual. The outcome five years later: the everyday citizen can’t even credit anymore the cost of the monthly subway pass! I guess we deserve it because that’s too rich right? going in crowded buses and packed metro wagons, god, the luxury is out of control! 😡 Punks.

[@CheeseypiPlays]

Voters DON'T vote for whatever is closest to their views -- They vote for whatever has been SOLD TO THEM as closest to their views. A strong social safety net would help no one more than impoverished rural voters, and if you talk to them couched in words blaming corporate greed & such they even agree with fairly strong regulations, but republican policies are sold as "increasing freedom by reducing government", so they vote against their own interests

[@justcommenting5117]

@@CheeseypiPlays talking to them blaming corporate greed is also selling them something as closest to their views.

[@MZero8099]

This seems way more complicated than what actual politicians do, which is just lie about what their policy does or who it helps using millions in marketing.

[@illbeyourmonster1959]

Also helps a lot to have a very large and diverse group of protected useful idiots to fight for the false narratives everywhere all at once. (The far left) Also, a good dose of willful cowardice in the majority that knows better that keeps them from fighting back in any way that matters helps a lot too. (The rational left, center and right)

[@valtersanches3124]

@@illbeyourmonster1959 while you're focusing on whose narrative is right, every politician, left or right wing, exploits you for all you have. That's their game

[@robinsekeris1721]

​@@illbeyourmonster1959How much current policy is far-left? I don't see any big influence from far left political agendas, could you please give examples of how "the far left" is having any impact on major institutions or states.

[@Tom-dn6zy]

​@@robinsekeris1721 you must live under a rock lol

[@DeepSeaLugia]

@@illbeyourmonster1959 didn’t know that not liking socialism is “far left “

[@comrade_marshal]

I had unknowingly used a crude form of the policy shifting method while leading a group projects few times and it worked favourably most of the time without anyone noticing it much. Now I see that there is a goddamn mathematics behind all such stuff

[@jimbrookhyser]

Sounds intriguing! Do you have time to write up any details?

[@EvilSapphireR]

@@jimbrookhyser the guy is basically saying he manipulated a few people once in a random group project.

[@gnaskar]

There is goddamn mathematics behind everything if you look hard enough.

[@simonspacek3670]

@@EvilSapphireR Well... who never manipulated few people? Think about all school projects, all games you played and all boring work meetings and tell me honestly if you never manipulated others. Plus, it is good to know how manipulation works, because then you will be less likely manipulated.

[@JoeARedHawk275]

@@simonspacek3670 As long as you don’t treat them merely as a means… I guess it depends on the severity. If it’s something inconsequential or leads to a mutual benefit, then it’s somewhat more acceptable.

[@sg-po5xq]

For those wondering why A wins over B or B wins over C in the first example. He assumes that each person has 1 vote in each vote and it plays out like this between A vs B: - the first person would pick its preference: A - the second person would picks its preference: B - the third person picks its preference between both options and that is: A Result A: 2 B: 1 Note only 1 vote per person, so person 1 can't second vote and chose its 2nd preference which is Policy B BUT this only works when this 3 persons with this arbitrary preferences vote. If you would have 3 persons and letting person 1 and 2 having the same preferances but changing the preferences of person 3 to: 1. Policy C 2. Policy B 3. Policy A the same vote between A and B is won by B. Having 3 options and arranging all of those 3 options in distinct orders has actually 6 variations. Its called permutation and the number of permuations is calculated by the factorial of all distinct options. This means 3 distinct options makes 6 different arrangements. 3! = 1*2*3 = 6 So in a world with more than 3 persons but perfectly balanced preferences he would have 3 more persons with the following preferences: Person 4 1. Policy C 2. Policy B 3. Policy A Person 5 1. Policy A 2. Policy C 3. Policy B Person 6 1. Policy B 2. Policy A 3. Policy C In his assumption he is only viewing this problem from half the possiblities

[@AnandBaburajan]

Thanks for the explanation! : )

[@carl-marvin]

Yes, thank you. I was just about to stop watching this 'nonsense' while he actually makes a valid point according to your explanation.

[@randompersonoftheinternet8012]

I noticed that and jumped to the comments. Thanks for the explanation. I think that this was still a very weak example though, especially to start out the video.

[@watchdognepal]

sorry i give up. i can't understand a thing

[@blackeyeole]

but this is not a true scenario at all. because the 3rd person should have the option to vote blank or not vote at all . so the problem is not an actual problem .

[@Thurston86]

The word “policy” has now lost all meaning but I feel like I might’ve learned something. 👍

[@archravenineteenseventeen]

Always has been. Watch rule for rulers in KZbin by cp grey

[@thesoundsmith]

It's now more cerebral. Cerebral Policy...😑

[@aarondcmedia9585]

It's called, "semantic satiation".

[@AByteofCode]

@@aarondcmedia9585 beat me to it by 4 hours on a 2 year video :/

[@aarondcmedia9585]

@@AByteofCode woooo!

[@z.m.6053]

Another thing to notice is that, if you look at how the policies shift over time, almost every voter is eventually subjected to a policy that makes it much worse for them. It actually resembles a kind of war. There’s a lot of back and forth and the path towards the red policy isn’t clean. It goes to show how such a process really hurts everyone involved.

[@colinmarshall6634]

The model pretty much describes the past 30 years in the US too. We went from marginal differences to wild extremes, just like this model did.

[@z.m.6053]

@@colinmarshall6634 fr it’s spin the bottle, and whoever the bottle lands on is the minority that’s going to be tortured by the ruling coalition.

[@neurofiedyamato8763]

​@@colinmarshall6634 exactly what I thought. It depicted partisanship so well. Nowadays we are going to shit because of it

[@bishopp14]

​@@neurofiedyamato8763 I just shit but I think it was because yesterday was taco Tuesday.

[@Shifftee]

This model describes capitalist “democracy”, and since capitalism means permanent competition (a war between everyone) the “democratic” process naturally reflects the economic part of the society. If we take Soviet democracy, there were no such flaws since everyone elected their own delegates from their own councils which they could recall at any given time. Here’s an excerpt from Stalin’s election speech: “ Further, comrades, I would like to give you some advice, the advice of a candidate to his electors. If you take capitalist countries you will find that peculiar, I would say, rather strange relations exist there between deputies and voters. As long as the elections are in progress, the deputies flirt with the electors, fawn on them, swear fidelity and make heaps of promises of every kind. It would appear that the deputies are completely dependent on the electors. As soon as the elections are over, and the candidates have become deputies, relations undergo a radical change. Instead of the deputies being dependent on the electors, they become entirely independent. For four or five years, that is, until the next elections, the deputy feels quite free, independent of the people, of his electors. He may pass from one camp to another, he may turn from the right road to the wrong road, he may even become entangled in machinations of a not altogether desirable character, he may turn as many somersaults as he likes-he is independent. Can such relations be regarded as normal? By no means, comrades. This circumstance was taken into consideration by our Constitution and it made it a law that electors have the right to recall their deputies before the expiration of their term of office if they begin to play monkey tricks, if they turn off the road, or if they forget that they are dependent on the people, on the electors. This is a wonderful law, comrades. A deputy should know that he is the servant of the people, their emissary in the Supreme Soviet, and he must follow the line laid down in the mandate given him by the people. If he turns off the road, the electors. are entitled to demand new elections, and as to the deputy who turned off the road, they have the right to blackball him. (Laughter and applause.) This is a wonderful law. My advice, the advice of a candidate to his electors, is that they remember this electors' right, the right to recall deputies before the expiration of their term of office, that they keep an eye on their deputies, control them and, if they should take it into their heads to turn off the right road, get rid of them and demand new elections. The government is obliged to appoint new elections. My advice is to remember this law and to take advantage of it should need arise.”

[@shta2]

I knew about the ABC example already but the demonstration of how you can get anywhere in a 2D place by chosing the vote was fascinating!!

[@johnchessant3012]

This is very interesting! "Don't vote for marginal gains for ourselves when it means making others in our society much worse off" is excellent advice :)

[@Berelore]

Na, it's ignorant drivel that superficially appears deep.

[@jimbrookhyser]

This conclusion was the part of the video I had the hardest time with. I think mostly because the idea wasn't sufficiently fleshed out. From what I've learned people are VERY likely to vote against their own self-interest because they DO believe in something being "better" or "more right" than what would benefit them directly. This prosocial outlook is common among us humans. That said, whatever I think is better for all, is still MY preference. What's missing is a reliable way of telling when my preference really IS taking others into account or not. I hope such a thing can be found, but I'm afraid that the best solution we've found to this problem so far is... well... voting itself. 😒

[@Fred-yq3fs]

@@jimbrookhyser Ppl being less politically aware and more self-centered make em easier preys. Segment audiences, emphasize the gains for each target so they decide for change -however symbolic/marginal, and you'll move the whole system to where ever you want. Audiences used to be less segmented / segment-able as they are today, so it was more difficult to pull off.

[@sillymesilly]

@@Fred-yq3fs how countries get invaded. Especially British took advantage of this.

[@soheil5710]

​​@@Fred-yq3fs This has me believe the strong political divide in America won't be resolved anytime soon. It's just too beneficial to the elite.

[@Puzomor]

For the people not getting the first part where A/B/C preference is explained: The premise is that there are ever only two policies being chosen from. Voters can never choose between all three. Voters can only choose between two policies at the time. If you pause at 1:16, you'll see that in case of A vs B, two out of three voters prefer A to B, and therefore A would win the vote against B. Same thing can be said for B vs C. Key takeaway is that the same thing can be said for C vs A - two out of three voters prefer C to A and therefore C would win the vote if it went against A. The video could have definitely explained it better.

[@jkw203]

Thank you. I was quite confused by it.

[@timothygooding9544]

the intro is just straight up incorrect. if the first 2 choices are being counted, all the votes tied, in each example

[@nifty7988]

@@timothygooding9544 Only one vote is being counted per person. Thats where it's confusing people. The "first, second, third choice" example in the video makes it seem like a "ranked choice" system is being used, e.g. each of the voters choices are being weighted according to where they ranked them, but thats not what he was meaning to say. If the vote were strictly A vs C, person 2 and 3 both have A listed below C, so both of their single votes would be for C, while person 1's vote would go to A.

[@Puzomor]

@@timothygooding9544 as nifty said, the ranking is probably the most unfortunately confusing part of the video, because it's not a ranking that a voter made. The voters never vote for all three policies, and never assemble a list ranking all three policies. The voters are only ever aware of two policies at the time, and the ranking list of three policies assigned to each voter is something only we as viewers know. The voters don't have a list and it's actually important that they only know about two policies at a time, otherwise this doesn't work. To properly use the lists you must pretend the third option literally doesn't exist while the first two options are voted for. This produces a situation where there's always two voters whose first choice is let's say A, and their second choice is B while the other only one voter has it the other way around.

[@SJrad]

I’ll dumb it down: It’s basically rock paper scissors

[@maxmuss4969]

Two takeaways: 1. As voters we need to be more cognizant of agendas and those setting them 2. As someone with an agenda (at the end of the day, all humans have an agenda even if it isn’t something nefarious), this is a really interesting tool for social manipulation

[@MrWesford]

But, why vote?

[@IceMetalPunk]

You missed an important takeaway. It's not just that voters need to be more cognizant of agendas and agenda-setters, but also that voters need to be more willing to listen to what other people need in addition to their own preferences.

[@madeline6951]

@@MrWesford because voting is one of the few ways to exsert power over your circumstances

[@yousufzohair4342]

Best Take away Democracy and capitalism shouldnot be the way of the world

[@madeline6951]

@@yousufzohair4342 wait what Democracy is good, we just have to monitor who has the say over the agenda. Which makes capitalism the issue.

[@Darth_Insidious]

Reminds me of how I used similar practices to create horrible dystopias in Democracy 3.

[@jonahbenjamin4496]

Very interesting Brian, this needs to be more popular! Just don't let lawmakers see it...

[@luiza4142]

i think the same...

[@yash1152]

do u really think they dont know?? wow.

[@thesoundsmith]

They've been DOING it since Reagan if not before.

[@andrasfogarasi5014]

@@thesoundsmith They've been doing it since the Roman Republic

[@silenthawkstudios9924]

@@andrasfogarasi5014 Greek* (or probably whatever "democracy" was first.)

[@raul_jocson_]

Thanks. This is one of the most eye-opening videos I've ever seen. It explains why policies can sometimes be seen to jump from one dimensional extreme to another, analogous to prying open a hole by pushing at its sides. Using the budget example, if a budget is already far below a voter's preference, then flipping it to far above their preference won't matter to them because in their mind "the budget is already screwed up". But if there's a simultaneous small shift closer to a different dimensional preference, then that voter will vote for that change thinking "at least we've won that small battle, the budget can be fixed later". Scary to this this play out in the real world.

[@amanyasir]

In India voter have 1 vote and each vote have value=1 so we can either choose a,b or c not and not all of them So does that make our system transative ?

[@nikelsad]

@@amanyasir no, in the vid voters chose 1 option between 2. It's a personal preferences what break transitivity.

[@Plato6969]

you're scary

[@KillerTuber32]

Ah yes, totally equal extremes like letting trans people use bathrooms vs making it illegal to be trans, labeling it sexual child abuse for you to dress in clothes the government has deemed for the opposite sex and making that crime punishable by death. Very equal extremes.

[@tobenweston1687]

This is why Australia has a preferential voting system. ‘I prefer this policy, but I would rather this other policy a bit further away from me over something extreme’

[@camt8804]

@@tcideh4929 That's right it's not perfect so it must be worse than what you have :D

[@camt8804]

@@tcideh4929 It's more just a comment on how you responded.

[@yousufzohair4342]

​@@tcideh4929 Well If it is Democracy then it won't be Good.. Democracy is a pretty rubbish form of Government.. Imagine Taking Opinion of people who have no Expertise or Knowledge of that topic over an Expert's opinion... It is like asking Normal people should we add water to Sodium sand to prevent it from catching fire .. Every person who has no Knowledge of chemistry will Add water to that sand because it is what he thinks right..And that wouldn't be even close to the right decision..So Yeah..There is no way Democracy can be good over the Right Governing Systems

[@yokowan]

I absolutely love this analysis! it's a really simple way of explaining a phenomenon that really dominates modern democracies. sadly, the hopeful note of "not acting only in self interest and cooperating" only works insofar as people can agree on what the greater good is. the personal values of individuals vary so much, that it can be very easy for one voter to think that a policy is beneficial for everyone, while another voter thinks it is harmful for everyone. both voters are acting selflessly, in the belief that their choice is for the good of everyone.

[@velraven8944]

The solution has been known and staring everyone in the face for over a century, but we've been made to forget it. Liberalism is foundationally based on the harm principle; "Liberty consists in the freedom to do everything which injures no one else; hence the exercise of the natural rights of each man has no limits except those which assure to the other members of the society the enjoyment of the same rights." The very core of the democratic society was an underlying value system of accepting that as long as you aren't preventing others from living as they wish, you are allowed to live as you wish. It's not inaccurate to say that "live and let live" is a deeply tangible philosophy and not just a catchphrase. This is the "not acting in self-interest and cooperating" because, as pointed out in liberal and economic philosophy, while it may seem counterintuitive, it is possible to act in one's own self-interest BY cooperating with others. The individual values of each citizen can vary as wildly as they wish, so long as the single underpinning value that unites the national identity remains "my rights are your rights". It becomes easy to manipulate a system based on freedom only after you demoralize the people into no longer holding Liberty and Justice for All as a value, effectively pulling the rug out from under that society.

[@memesterjohnson4096]

​@Velraven this breaks down when your rights are not their rights. we are still arguing about who rights are superior to each other. look at hate speech laws as a example. the words don't physically hurt any one but because some one made you feel bad the government want to solve to problem by saying this speech is bad.

[@nicholaslowry2366]

@@memesterjohnson4096 I think what you’re missing in that example is that hate speech can lead to and encourage hate crimes. By making a group to be less than human, you can make the more gullible of society believe that they don’t deserve the same protections as a human. This is a tactic the Nazi party used to make the ‘final solution’ seem more palatable to the German people.

[@teebob21]

So what he's saying is that in order to best serve society, voters need to vote against their own individual best interests. HOLY SHIT Republican voters have been doing this for years and we've been laughing at them for it!

[@teebob21]

@@velraven8944 dat libertarianism doe

[@ascrassin]

one thing to note is that some of those problem can be diminished by using other type of voting system. mainly by creating a score system to take into account the seconds choice (like asking in a scale of 0 to 10 how much do you support a proposition).

[@mlafleurhua]

Very interesting video. In some sense this is a mathematical model of the political process called « shifting the overton window » in the US.

[@digaddog6099]

The Overton window is about where most people are, not where the policies are.

[@tennicksalvarez9079]

@@digaddog6099 u might be able to blend the two ideas idk

[@waynebimmel6784]

​@@digaddog6099 It is more about what is considered extreme but tolerable by the majority. Marjorie Taylor-Greene level of racism is barely inside the overton window, most people know someone personally who has the same opinions. Kanye however...

[@Spellweaver5]

You misapply the idea of the Overton's window entirely. The window itself is an evident fact. You can't get political support if your position is radical. The idea that it can be shifted intentionally, rather than naturally, let alone the way to do it, are all in the realm of speculation, bordering on conspiracy theories.

[@ArmoredZephyr]

​@@waynebimmel6784 Lmao, someone who believed the whole JEwISh spACe lAzerS lies told about her.

[@SignificantNumberOfBeavers]

There's an interesting idea called Condorcet's Theorem that has a lot of implications for democracy. It states that if you have a collection of decision makers that are all at least slightly better than random, the more decision makers you have the more accurate the overall decision is. It's what makes machine learning algorithms like random forests work.

[@pentscoff6969]

That's basically what is being exploited in this model

[@remoosecode7558]

What if they're slightly worse than random?

[@1ucasvb]

It's not very relevant for democracy. Condorcet's Jury Theorem only makes sense in binary 1-D decisions, and only if there's a globally "correct option" and agents are noisy samplers of that. (Hence, the framing of a "jury" choosing guilty/not guilty). It doesn't apply much to democracy in the real world, as people do fundamentally disagree on what's "correct", and most decisions involve judgements on multiple attributes, not just one.

[@alistair676]

If I'm not mistaken this applies to the financial markets and prices of assets too, where with enough people trading, the price will be remarkably accurate.

[@DodgeThatAttack]

​​​@@1ucasvb interesting to note that a 2 party system is basically a binary decision. The idea of correct or incorrect isn't actually a necessary, you can easily replace it with 2 abstract ideals such as a left and a right, and have people vote which is correct between them. The 2 party system itself disguised by the illusion that other parties CAN run, but they never win, so they are almost irrelevant to the result - in fact, having a 3rd party who's policies align more with 1 side can hinder them both as they divide the votes between each party with similar policies. Really the theorem is basic statistics- the higher the sample size, the mkre accurate your result will be. so in this sense, its totally applicable, even in a non binary scenario

[@petermoore3439]

You won my subscription with this single video. It'll be super useful for the classes I teach on policy analysis. Many thanks!

[@gf4913]

You built a cool model but this chain of proposals to the bottom can be easily blocked if you give to the people in the game the ability to make proposals. When a policy get too far away in the plane someone (probably the one more distant to it) would call for a vote, bringing a more popular policy back. BTW this is why the power to propose legislations is very important and in every democratic system each MPs, in addition to voting bills into laws, they can propose new ones.

[@dullyvampir83]

You just need to corrupt the MP's, very easy for powerful interest since there are not that many and they are reliant on money for their campaigns. Unless their are public initiative referendums the model above aplies very well.

[@patriciusvunkempen102]

yeah but how would that work with more than a few hundred people? and then you have to attest the fact that mos tpeople aren't that smart and then a lot people arent that interested, and then people also have delisions and other insterests. and it wont work. look at the real world. majoritys in western nations crassly oppose mass immigrations, there are big majoritys there and its just ignored bc they push other issues to the forefront. and then propagate other issues as more pressing, another crisis every half year

[@gf4913]

@@dullyvampir83 Actually the above model apply more to public referendums rather than a parliament. Public referendums are often called by small group of activists and then the population at large is called for a vote on a few selected choices. I can see how a small interest group calling for the referendum could be associated to the agenda setter present in the model.

[@dullyvampir83]

@@gf4913 True, but it would be lot harder to control the agenda compared to having a bunch of politicians setting the agenda alone.

[@mattpattok3837]

This model doesn’t only have to be about particular policies. In the case of a representative system such as the US you can apply this to politicians, and the model fits well. Politicians get elected not for their fit with their constituents, but because about half the population thinks they’re slightly better than the last politician. So over time our politicians get farther removed from anything anyone wants.

[@1ucasvb]

Very good explanation of the theorem. This theorem is also a great way to highlight that while majority rule based on ranked preferences seems intuitively "correct", it's actually arbitrarily anti-democratic and anti-representative. But it's not a problem with democracy, it's a problem with ordinalism (the idea that only ranked preferences matter) and majoritarianism (the majority of preference wins). Using cardinal consensus-based collective decision method, this issue doesn't exist as the intensity of support/rejection can factor in explicitly, and the decision is stable. It also actually captures the intuitive notion of "representing a majority" that most people think about.

[@mikip3242]

What is "cardinal consensus-based collective decision method"?

[@thebiscuitguy646]

So you're saying, instead of ranking policies best to worst, rank policies on a scale of 1-100 and whichever policy has the highest total after voting is done gets implemented?

[@1ucasvb]

@@thebiscuitguy646 Yes, but that's just one way of doing it (called score voting), and 0-5 or 0-10 is enough. But to be precise, the correct word there is "ratings on a scale". Ratings are a form of cardinal evaluation, which capture ranking information as well (preferences), but they also contain more information, in particular about degrees of indifference. (More on that on the next post.)

[@1ucasvb]

​@@mikip3242 Here's a detailed explanation. The lesson from the video/theorem is that assuming a majority of '"slightly better" overrides "much worse" can lead to arbitrarily bad decision-making, if applied iteratively. There are two types of information you can capture from voters: ordinal (rankings, orders of preference) and cardinal (numbers, ratings, degrees of evaluation, strength of support). You can deduce rankings from ratings, but not the other way around. This video is about how to make collective decisions using rankings only. Once you have collected the information, you aggregate it to make a collective decision. There's many ways to do it, but they are constrained by the information you collected. Regardless of how the decision is made, democracy requires the assumption that all voters opinions are equally commensurable (can be measured on the same scale as one another), otherwise you cannot look at opinions in aggregate. Strictly from an ordinal (ranked) perspective, you can only justify making any decision through unanimity (if everyone prefers A>B, then you choose A over B), as you cannot justify violating any voter's B>A preference. This is called Pareto optimality, but it can be arbitrarily anti-democratic. To address this, ordinal voting methods must invoke majoritarianism: you allow a majority of preference to violate the preference of a minority, but this requires "counting", which is cardinal information. It's a rule of thumb we came up to capture a notion of "consensus" ("general agreement") under rankings, based on that equality/commensurability assumption. You can already see from this that decision-making needs to be inherently cardinal at some level. But even under majoritarianism, if you only collect ordinal/ranking information, you must also assume every A>B is equally strong in preference, with voters expressing infinite precision in each preference, and being 100% resolute (as if voters had zero uncertainty about the options and their own opinions), so degrees of indifference and similarity between options disappear. The video illustrates how this distortion ("slightly better" cancels "much worse", @ 4:45) can be amplified without bounds, because you're always turning "indifference" into "full-strength preference". Under cardinal methods, voters *can* express degrees of indifference between options, by simply giving options similar evaluations. One possible way to do this is by letting them give a rating (say, 0-10) on each option then choosing the one with greatest score (called "score voting"), but could also be conceptual like "Good", "OK", "Bad" (See 3-2-1 voting, for example) or simply "approve/not approve" ("approval voting"). The point is that "10 vs 9", or "Good vs OK" are in a fuzzy sense "similar", and different than "10 vs 0" or "Good vs Bad". Note that, in the case of scores, this is different than things like rating products, because every voter is evaluating all options together at the same time, and every option gets independently rated within that shared, comparative context. This difference is fundamental and many people miss it, and it's what sets the common scale across multiple people. (It's important to mention this as a lot of people get put off by "voting with numbers", but rankings make even stronger assumptions.) This shared context, and that basic assumption of commensurability under democracy, means that we assume every voter's best-to-worse scale is the same. Clearly, the ratings are fuzzy and imprecise, but this is a good thing. There is no precision in the real world, and consensus *requires* measuring how much options "overlap". We want "10 vs 9" to be "somewhat similar". This is why cardinal methods are "consensus-based". I say "decision-making" because this isn't just voting, the point is to collect and use the information about indifference when making group decisions. Since cardinal voting ballots existing within a comparative context (options aren't evaluated in a vacuum), in practice, this means voters compress their best-to-worst scale to emphasize the distinctions that really matter to them. The result of using a cardinal method here is that "slightly better" and "much worse" will not really cancel out, so it's easier for a "general agreement" (consensus) to emerge out of the picture as the winner, and the addition of "slightly better" or "much worse" options won't affect the results much. As simple illustrative examples, using the same scenario as the video, let's assume [Current Policy] seems "pretty good" for everyone, and [Red Fringe Policy] is the problematic one preferred by a majority, with [New Policy] the one you're trying to push for skewing things. Under *Score Voting*: 3 voters: [New Policy]=10 [Current Policy]=8 [Red Fringe Policy]=0 ([New Policy] > [Current Policy] is "slightly better") 2 voters: [New Policy]=2 [Current Policy]=8 [Red Fringe Policy]=0 ([New Policy] < [Current Policy] is "much worse") Total scores: [New Policy]=34 [Current Policy]=40 [Red Fringe Policy]=0 ([Current Policy] wins) Under *Approval Voting*: 3 voters approve: [New Policy] [Current Policy] 2 voters approve: [Current Policy] Total approvals: [New Policy]=3 [Current Policy]=5. ([Current Policy] wins) Adding any slightly similar policy is stable here, as evaluations are given independently under cardinal methods, so the new ones won't affect the other ones much unless they're *much* better, not just "slightly better".

[@icecreambone]

@@1ucasvb while I appreciate the detailed explanation, i think a summary is in order 1) ordinal voting (voting A > B or B > A) assumes that how much better or worse A or B is than the other is the same for everyone, when in reality it can be far better or worse for one group than another, and this leads to undemocratic things like concentration camps 2) cardinal voting tries to resolve this by letting people actually rate both A and B (and C and D etc. all at once) on a scale and then tallying up their scores now for some questions - would it be accurate to be concerned that sufficient polarization can turn cardinal voting into ordinal voting? in the approval voting example which this is most similar to, voting for both policies seems equivalent to an abstention in ordinal voting, so there isn't particularly anything stopping them from approving only [new policy] in an attempt to game the system even if they are okay with [current policy]. in an example with more than two choices and two groups, the ability to vote against one particular thing does seem to be an improvement over ordinal voting, but this can be circumvented by a similar process to the video, by only raising two policies for a vote at a time (on the flip side, you could argue that ordinal voting results in more polarized thinking to begin with since everything is a nail when all you have is a hammer, but in general i think it's good to ask how the system affects and is affected by public consciousness)

[@troybaxter]

I agree with your last point about how we should sometimes vote against our best interest, however, it is hard, nay near impossible, to tell if a policy is closer to the middle or an extreme that negatively impacts everyone.

[@lloydfromfar]

Wow, I was marginally convinced all along that this model has only deep impact on the world util the brilliant conclusion! Well done! :)

[@thatwellhungasian6925]

It's cool that the algorithm decided to pick this dude's videos. There's seemingly lot of new viewers to certain videos. Props to *you* Brian *Yu*

[@someoneyoumightknow4375]

I gave my dad one of the circle-drawing puzzles and proceeded to watch him struggle for an hour. finally, he asked me to solve it; it was only then that i realized the one i had created for him was physically impossible.

[@6-dpegasus425]

Hehe

[@SirZelean]

That's the thing. We have to get millions of people to cooperate to get things in our favor, while the ones who are working for an agenda are a few, organized, and motivated bunch. It's a much greater effort for us to get things our way. It's much harder to educate millions than it is to build a team of a few dozen people. They are bound to get things done their way, especially in larger countries.

[@Aim54Delta]

Not necessarily. For starters, this assumes all policies are equally valid from an operations standpoint or will be equally followed. There are also fractional movements. There's also the fact that the only realization that need be made is a simple one: "what is best as a systemic policy?" When I was talking to a lady at a book store about things - she admitted "you know, I have usually only voted for what was best for me, but after we have been talking, I realize that we need to vote for what we think is better for an institution." For example, there is a difference between the voter who votes for a policy that promises things they want and a voter who votes for a policy that functions the way they think better serves its purpose. An example of this is the recent legalization of marijuana bill in missouri. It passed, but not by my vote. Why? Well, for starters, I think pot smoking is annoying. That said, I do actually think the criminalization of drugs is unconstitutional at the federal level and I don't see reason to criminalize it at the state level. Not all annoying things need be illegal. However, they attached a weird system to it. Rather than just striking all weed infractions from law... they had a lottery system for who could become growers and distributors that would invariably become gamed by the friends of politicians for warlord politics. Had it been a straight decriminalization, I'd have supported it. What generally happens during these implosions of collective decision making is the rise of heroes and nobles - people who become icons of mutually agreeable positions as policies drift further and further to the absurd until there is a seizmic shift that is a revolution in its own right. Fundamentally, the agenda setters don't understand where these people come from. They respond with trying to have these people removed or shamed from society - and it doesn't work. New ones pop up, saying the same things - and eventually, the revolution succeeds.

[@The_True_Mx_Pink]

Which is why the dissipation of classes as a whole, or at the very least, oligarchs and synonymous, such as the 1%, has been considered a mandatory step in the right direction by millions, myself included, for a century. Classes turned Russia into a fascist dystopia, first with the USSR's ruling class and now with the Republic's oligarchies, and turned the US from a democracy to a plutocracy in all but name.

[@nickthompson1812]

And then give those few dozen people millions of dollars in resources to lobby and advertise… it’s no wonder corporations run our government

[@redrumthebum]

​@@The_True_Mx_Pink based

[@The_True_Mx_Pink]

@@redrumthebum Thank you. I like your username.

[@ohauss]

Mathematically an interesting concept, but realistically, it suffers from a couple of problems. a) Voters areas of preference will not necessarily be circular/spherical/etc. - they may be more willing to compromise on some issues than others. Which these are may vary from voter to voter. b) Repeatedly voting on the same topic within a short period of time will likely affect willingness to compromise and shrink the geometric distance acceptable - making it that much harder to find an overlap. c) As we've seen over the past years, with the right tools, it may well be easier to shift what the voters want (or rather, a large enough part of the electorate) by targeting susceptible groups with disinformation.

[@hastyscorpion]

A) The circle isn’t specifically important. You can draw what ever shape you want and this kind of thought exercise will hold true. B) The time period is not specified for this this could absolutely take place over a long period of time. C) isn’t a problem with this thought experiment. It’s a different subject entirely. Your critique suffers from taking the model far too literally and not going for the general principle.

[@ohauss]

@@hastyscorpion Your critique suffers from abolishing the very concept of a principle in favor of "anything goes", The shape is very much important, as shapes can be non-overlapping. The time period is also important, as larger periods of time means that the shapes do not necessarily stay what they were. There's a huge difference between using a general principle and keeping certain aspects fixed to simply a situation on the one hand vs. looking at something that is effectively a consequence of the model design in the first place. Know the difference between an outcome and an artifact.

[@hastyscorpion]

No… it’s not abolishing the concept of a principle. It’s saying that poking holes in a very broad strokes model because it doesn’t account for every specific scenario isn’t helpful or teach you anything. Yes the shapes “ could “ change over time, but they often don’t. Yes , the shapes “might” not over lap , but they often do. (also the shapes not staying where they were is not relevant unless you are have a specific point in mind that you are trying to get to which is not the point of the video. as long as there is overlap it’s still relevant) none of the problems you brought up indicate that the example is simply a product of the assumptions model and couldn’t actually happen. Because none of them say that the assumption couldn’t be true only that exceptions exist.

[@ano_nym]

Strangely enough c) only seem to happen when it's regarding something the one talking/writing about it disagrees with.

[@AkiraNakamoto]

You should take a probabilistic view. At least the videomaker has proven that there is a non-trivial chance that a simple form of voting goes very wrong.

[@cyb3r._.]

there were no captions for this video, so I listened with sound, and at the end of the video, I expected the classic Ted-ED outro music and only remembered this wasn't Ted-ED when it didn't play lmao

[@squiddler7731]

The main point focused on here is how voting only on your own selfish interests pushes towards something that's awful for everyone, but I think it's also worth nothing how the entire process only works because voters were only allowed to vote between two policies. If policy 1 was always on the table, people would've voted for it over the one worse for everyone. But because they were always restricted to two options (either artificially by removing all others or naturally by using a voting system that always tends towards two parties), people would always find themselves chosing what they believe to be the lesser of two evils instead of anything resembling what they actually want. That first example should've ended in a tie between all three policies, but the moment one of them became unavailable or unviable it allowed for the cycle of shifting the goalposts to begin

[@mindbreaker1337]

if this method would be applied in the real life, the voters would not know the third option until they already voted for the better seeming second one

[@squiddler7731]

@@mindbreaker1337 Not necessarily. Third parties do exist and people are aware that they exist, but no one votes for them because it's essentially seen as throwing away your vote. That's what I meant by an option being unviable: If you have exactly one vote and you're given three options to vote for (two really popular options and one that you think is better but is far less popular), then if your smart you'll never vote for that third option no matter how much you like it because you already know it will never win against the first two. Have everyone think like this and you get a two party system where no third party can ever win no matter how promising they seem to be. There's voting systems that can address this issue by letting you vote for your second and third choice as well (if your first choice is eliminated from the race it counts for second choice so that you don't have to worry about throwing away your vote), but as far as I'm aware few if any states in the US use this system.

[@rocklee5231]

The first example you gave was a tie in all cases if you use rank choice voting. Each option A, B, and C had a rank of 1, 2, and 3. This gave a net score of 6 for each policy. A=B=C.

[@ahsnsb]

I'm so confused. This video is up for two years. He would've noticed and corrected it by now if it were wrong.

[@reynoldfroese8439]

@@ahsnsb I doubt it. We exist in a societal context that rewards confidence and punishes integrity.

[@dinosaurusrex1482]

@Ahsan Shah the video does not use ranked choice voting, and neither do most countries

[@jimbrookhyser]

It's not wrong, it just considered the effect of preference and voting systems. This wasn't an algebra lesson. One could construct a voting system where voters state preferences and then those preferences are translated into points and then those points result in an even three way tie you described, but that's just not the type of voting that was considered. I am curious if any governing body uses a system like that. Bonus points if it's a democracy or representative democracy. I have vague impressions that systems like this might be used in sports team rankings, but I wouldn't know the details. Certainly worth thinking about.

[@auxilyatg5321]

​@@Preetzole except that in the B vs C case, both are equal with the both having a person with either policy as their middle favorite, as well as a person with either policy as their favorite. To say B wins is wrong

[@alittax]

0:54 I was first confused as to why that is the case, maybe someone else is as well so I'll explain. If it came to a vote, the first person would vote for A (because A is his first choice), and the third person would also vote for A (because he prefers A over B). Only the second person would vote for B, because that's his first choice. So 2 out of the 3 people would vote for A, and 1 out of the 3 people would vote for B, which means A would win that vote with a 2/3 majority.

[@twanzuidema3537]

Thanks, this really helped!

[@imdum21]

thanks, I posted a question in that regard. So yeah, you were right in assuming that someone else may be confused as well :)

[@alittax]

@@imdum21 Nice to hear that, and I'm glad I helped! Have a nice day.

[@alittax]

@@twanzuidema3537 Nice to hear that, and I'm glad I helped! Have a nice day.

[@neveklund3267]

It sounds like Arrows Paradox, that was taught in economics in the 1970's and first developed by economists such as Kenneth Arrow, and others back in the 40's and 50's. It operates because voting is done between only two choices at time, with a series of revoting with different choices. Where voting is done with several choices, and allowing a mechanism for narrowing in the range of choice on a second round of balloting or re-counting after dropping some choices the paradox can be over-come.

[@SyntekkTeam]

Great Video! I think the fundamental problem here is a choose one plurality system. If voters are only getting 2 choices and they can only pick 1 without expressing their level of support, then the more polarizing policies get an edge. This is solved if voters get many choices, and they get to express their level of support. That's why I like STAR Voting :)

[@crazydragy4233]

I also feel like the video doesn't address that policies are always about trading resources, and some positions are about hurting specific voters...

[@billyswong]

I think the issue illustrated here is far less about voting system, but far more about public communication, open discussion and be considerate to minority needs and opinions. If you are voting for who to sit in the office, of course you can invent a lot of ways. But for bills, there are always only approval-disapproval at the end. The only way to prevent malicious agenda setters using the radicalization skill to achieve things nobody wants, involves: 1) Hold an open and uncensored public discussion and communication so that everyone is aware if a new proposal will hurt some people a lot in an unjust manner while benefiting only very little to the general public. 2) Realize refusing to respect the minority will ends up ruining the society as a whole because everyone of us is ultimately a minority in some policy aspect. 3) Do as what said in the video, make decision not only based on one's own interest, but also take care of whether the current policy is hurting the minority too badly, and whether the proposed change will do so.

[@grantcivyt]

​@@billyswongWhat do you think about not putting things up for a vote in the first place? If you don't let politicians curb your constitutional speech rights, it's much harder to end up with a Chinese censorship regime through voting games.

[@billyswong]

@@grantcivyt Well, politicians have invented many tools to silence people. "Discrimination", "Hate speech", "Racism", "Wokism", "Fascism", "Communism", all sort of terms to numb people's mind into deaf ears.

[@zeluis9219]

Maybe the most remarkable thing posted on YT. An extraordinary piece of wisdom about things we never think of. AND AN ALERT about how to overturn democracies. Great stuff here! Thx

[@DukePaprikar]

0:50 What am I missing here? Two of the three voters prefer policy B as well (likewise for next steps; each policy is exactly once in each of the three places), so I don't see how any one policy 'wins' over any other. They all seem equally favored to me. Edit: nvm, got it.

[@janpavlu5547]

I haven't got it, still makes no sense to me

[@josephsylve6758]

​@@janpavlu5547 Remember that each people can only choose one policy. The policy on top is the one choosen.

[@DukePaprikar]

@@janpavlu5547 I'll explain it on one example, and the rest is the same thinking: look at 0:52; The first voter prefers A, the second voter prefers B, so they're equal up until now; but look at the third voter: they prefer A over B, so A wins. The reason for this is that when B is at the second spot, it doesn't count because A is at the first spot with that voter, but when A is at the second spot, it counts because there the B is at the third spot; the fact that C is at the first spot here makes no difference because they're choosing only between A and B here. It's analogue to this in the other situations.

[@reyne2077]

@@janpavlu5547 There are three posible policies, but electors vote only between two of them at time. So if your favorite option isnt on the vote now at all, you would have to vote for your second favorite one for now. So in the end not those who vote decide which policy wins, but those who choose which policies would be on vote now and in which order.

[@thoughtsfocused2262]

The argument that A is preferred over B is the same as saying rock is preferred over scissors after removing paper from the rock paper scissors scenario.

[@Ms_d93]

Best video I’ve watched. Keep going

[@alexleung842]

Would love to see a mathematical explainer on whether ranked choice voting solves this or not.

[@1ucasvb]

It does not, it actually makes it way worse. The problem underlying this result is that rankings and majoritarianism do not capture enough information to define a consensus over a multi-dimensional decision space (each A>B ranking necessarily collapses everything into a one-dimensional axis between options A vs B). RCV is even worse at this, because it is explicitly an anti-consensus method: at any elimination step, voters only support their current #1 option, and nothing else. Therefore, every "faction" is always supporting only their current favorite at very step, and there's no room for mutual support, ideological overlap or agreements to emerge between groups, i.e., no consensus. Contrary to what many people think, the transfers don't help with this, because you already eliminated an option, and that option could have been the consensus. In fact, any time RCV *actually* behaves differently relative to other voting methods (ranked or rated), that's precisely what it is doing: eliminating the consensus. The consensus is not always the favorite option of the largest groups, which is what you'd need for this. Here's a very clear example of this. Three groups voting on ABC candidates: 40% Love A, like B, hate C (A>B>C) 35% Love C, like B, hate A (C>B>A) 25% Love B, like A, hate C (B>A>C) B>A according to 60% of the population, and B>C according to 65%, so B is the "Condorcet winner", the candidate who beats everyone else in 1-on-1 preferences. This is the best approximation of a "Consensus" you can get under rankings, but it's very flawed too (see below). In this example, we can see that B is either "loved or liked" by 100% of the voters, so it's actually a great choice. But this analysis requires us to look at all opinions "at the same time", which RCV does not do, and beyond just the rankings. RCV does not ever take into account the A voters's support towards B as their #2 favorite. Only favoritism matters under RCV, and favoritism is not the same thing as support. So the result under RCV is that B, the consensus candidate universally "liked or loved", is the first eliminated. You then have A vs C only, and C wins with 60%. RCV worked as intended here: it eliminated the consensus, and chose the dominant in the absence of a consensus. This is why RCV generally promotes factionalism and bipolarization, ultimately, contrary to what most people claim. it only makes sense if all options ultimately collapse into two mutually-opposing factions. The above is about single-winner RCV (Instant Runoff Voting). The multi-winner version, Sngle Transferable Vote (STV), suffers from similar issues but it's more subtle, but it also deviates from the subject of the video. Most people's claims about RCV being good for "third parties" are mixing up the single-winner vs multi-winner versions of the method. As promised, here's how rankings are not enough to capture consensus. The issue is that ">" does not distinguish between "Love > Like", "Like > Hate" and "Love > Hate". They're all deemed equal, which is a huge distortion, and what leads to the problems in the video. Consider instead a scenario where people are leaning more towards disagreement: 40% Love A, hate B, really hate C (A>B>C) 35% Love C, hate B, really hate A (C>B>A) 25% Love B, hate A, really hate C (B>A>C) The rankings are exactly the same as before, so B is still the Condorcet winner, the best "approximate consensus" you can get from rankings. But clearly, B is not a great candidate either, as it's only positively received by the smallest faction (25%). Since there's no consensus under this scenario, you're better off favoring the largest faction you can find between A or C, and in this case at least A is loved by 40%. You can't positively satisfy more than 40% of voters, so that's the ideal pick. If your goal with voting/democracy is to "positively satisfy the largest number of people as possible", rankings are not good enough. Note that, under this scenario, RCV works as intended. You eliminate the "pseudo-consensus" of rankings, and you focus on favoring the dominant faction, which in this case is A. This exemplifies what I just said. If there is a real consensus, then if RCV behaves differently than any other voting method it would be by eliminating it. If there is no consensus, then it would find the most dominant faction to favor. The point of RCV is to pretend there's never a consensus to begin with, so that it always strives for finding a dominant faction. It effectively assumes that every election is as polarized and antagonistic as possible. This is why I like to say that RCV is a conflict resolution method, not a voting method, so it shouldn't be used in politics. Choices made under democracy should represent everyone as much as possible, not simply favor the dominant faction.

[@DoctorCyan]

@@1ucasvb Hell yeah dude, I love the essays you're writing in here. Score voting forever!

[@jerel42]

I'm a huge RCV fan, and a huge RCV activist, and RCV does *not* solve this problem. However, the scenario set up here, where they hold election after election on a single policy, is not at all realistic to the real world, rather it was to illustrate a point that "agenda setters" hold enormous power. They **do** hold enormous power, but it isn't used like that example in real life. No voting system can be perfect, so really you need to *compare* voting systems, not just look at them one at a time. If you assume voters will vote sincerely -- not strategically -- then lots of systems work really well in public elections. However, huge numbers of voters vote strategically in practice, often encouraged to do so by campaigns. RCV is incredibly resistent to strategic voting (it's "NP-Complete"), especially elections that have lots of votes such as public elections. And winners require both enthusiastic support and broad support -- that's a core value of RCV. Therefore, in the real world, for very high-stakes and emotional situations with lots of voters -- like public elections -- RCV is almost always the best choice. Also RCV has a very long track record in the USA and internationally, so we know empirically that it tends to work very well. For those familiar with the Condorcet criteria: RCV and Condorcet are extremely close to each other in real world usage. In the USA, when there is a Condorcet winner, which is the vast majority of the time, 95-99% of the time RCV and Condorcet pick the exact same winner. And on the rare occasion when they don't coincide the RCV winner is still a very valid winner, because the Condorcet winner didn't have much enthusiastic support. A shout out to Condorcet, Range, STAR, Borda, Approval, and Negative Voting -- they are all valid and top choices in certain situations, and I personally have used or recommended almost all of them at one time or another. But for public elections, RCV is my clear #1 choice.

[@jerel42]

Oh, I just had another thought -- let's say there are five policy proposals for say, how to use some land in a city: proposals A, B, C, D, and E. If you put up all five for an RCV vote at once, it would be the opposite of the horrible manipulation shown above. The winner would be the one with enough enthusiastic support to stay in the race, and enough broad support to win the race. In that sense, if that is what you mean, RCV *would* solve this problem.

[@mmarinete1116]

Score Voting, Approval Voting, Single Transferable Voting and Condircet Method

[@6piyushc]

Very informative video. A couple of things: 1. A policy that's suboptimal for every voter must be optimal for policymakers on some dimension, which could be the same dimensions that matter to voters or different. What could those dimensions be? 2. Even in the real world with many many dimensions, most problems can be broken down into 2 or 3 dimensional spaces. If the voters have already started voting for suboptimal policies, how do we bring them back?

[@OGBIDENCUSH]

lead, lots and lots of lead

[@davidlewis6728]

policymakers like policies that give them more control. the only proven way to humble an unaccountable policymaker is with lead.

[@anastasiaklyuch2746]

Another option is the strength of secondary choices: In "3 best" voting, set 1st choise to 1 vote, 2nd choice to 2/3 vote, 3rd to 1/3 vote (each person has 2 votes to spend thus) This way the sum of all votes in the first scenario will be equal between options A, B and C. Vote strength must always be equalized this way is multiple choices are allowed

[@Drone256]

This is why we elect representatives and we are not a democracy.

[@4dtoaster819]

You can think about the green area as how many options the policy makers have, or in other words how much power they have. This model then showes how policy makers increase there power by increasing discontent. They did not need to gradually move the policy closer to their desired point. It was even advantageous to move the policy father away from their goal, as long as it increased discontent.

[@JackBond1234]

Your point at the end is very salient. Democracy doesn't work because you have a prisoners dilemma, hoping people will cooperate rather than harm others out of self interest. That's why it's important to define rights that are off limits no matter what. Then people have to change what words mean in order to harm others out of self interest.

[@l9324]

John Rawls makes a big point of this in his book Political Liberalism, that the ideal society should enumerate the most essential "basic liberties" as unalienable (with respect to everything but other basic liberties, which would have orders of importance), then work off of those as a universally agreed upon point of reference. it's a lot more coherent and intriguing than i've explained, so if you're at all interested in lines of thinking related to this i'd 100% give it a read

[@MAlanThomasII]

My observations of chronically-unstable political situations is that, over shorter time scales, you often end up with voter exhaustion, where they just refuse to keep voting / changing things and vote for whatever provides the greatest stability for a time. Over longer time scales, the agenda-setters become one of the things the voters change in the process of successive votes for change, so it's very difficult to pull off such a maneuver from the inside. Outside groups find it much easier to simply change voters' preferences, whether they be members of the general public or a legislature, than to try to set successive agendas from the outside. It's sort of like the Gibbard-Satterthwaite theorem; while tactical voting must be possible and is practiced, it's not always practiced to an extent that determines elections.

[@bartbartholomew]

You don't need to have an actual vote to do this though. You can adjust current opinions the same way with news cycles and such. Then when people go to vote, they vote where their current opinion is.

[@rdean150]

This is brilliant. Love it, love everything about it.

[@ralphmerkle9315]

Thanks!

[@BarcelonaMove]

I've seen that happening for years in Spain and is crazy to see it mathematically explained.

[@amegatron07]

Thanks for video, it gives a nice visual understanding of not only how any policy can be chosen by the society, but also how public opinions can be manipulated. Unfortunately, this has been a harsh problem in my country for the last 20 years. The whole world now sees the consequences. But there is still an important thing that has to pointed out. When speaking about democracy as a political regime, one should understand that democracy (as just a voting procedure) is nothing by itself, without the protection of human rights. Moreover, democracy is not more than just one of the implementations to realize human rights. Without them, there is no sense to be happy about free voting, because your choice is rigged by default. Or if we come back to this particular scenario in the video, people are only bound to chose from what they are suggested without the ability to propose their own policies, or at least publicly say, that the options they are given are bad at start, or that they may lead to bad consequences. So, summarizing, the core of a healthy society is not democracy by itself, but essential and irrevocable human rights. Democracy is just a direct consequence of them. If we need to protect the society from going the wrong way, we must assure that this foundation can not ever be altered in terms of relaxing this rules. And important thing here is that society itself must understand this foundation, not just the authorities they voted (or "voted") for.

[@naneri]

The true democracy was in Ancient Greece. The key principle of democracy was every citizen feeling his responsibility for others. Basically - everybody was a politician. Nowadays 99% are not interested in politics and make a surprised Pikachu face when they see what happens outside of their homes.

[@safetypinning]

aristotle said the best size for a city (assuming they are a democracy) was 5000 people. this was so people can act in accordance with the greater good while holding each other accountable for it. what aristotle said is definitely a criticism that holds validity in the face of modern democracy. how can we work together when we are such a huge population that is constantly being manipulated by policy makers and pitted against each other?

[@tonyisnotdead]

we have social media, much wider and accessible education and a system of government where it's not just the state vs the people

[@nphiro5030]

I watch this channel just as much as Numberphile. Well done.

[@thomaslindblom8926]

Great video! The concept your describing here seems more less like the principle of “Tyranny of the Majority”. The idea that the policy preferences of voters in minority factions have zero weight with respect to decided upon policies that become law. While this scenario can be prevented if people vote in a more altruistic fashion, the more realistic solution is to do away with a simple majority rules system.

[@wyjax0685]

This has occurred several times in American government, the great compromise allowed for the senate to protect smaller states while the house remained population proportionate. Also, the electoral college helps keep the presidency from tyranny of the majority.

[@davidlewis6728]

it's more like "tyranny of the majority" is inevitable for democracy, but policymakers who decide what the majority gets to vote on can manipulate it to turn it into a "tyranny of the system-setters". so not only do the minority decisions carry zero weight, but the only ones whose opinions matter are the ones who organize the elections and count the votes. neither scenario can be prevented if people vote altruistically. the only ones who know what's best for any given person is themselves, and they were going to vote for what they believed would benefit society as they knew it the most, anyways.

[@egilsandnes9637]

Very nice! Though it should be pointed out that even if all voters prefered policy is of the kind "This is what I think will make the best society for us all", they might still differ hugely*, and the mathematical principle shown here would still hold. I think most people actually believe that the policy they vote for is probably the best in general, not only for them personally. *That is assuming that each voters preferred policy in the most egoistic way will always be more spread than a policy each voter think would be best for all, but that might not always be the case. I think it would hold in most cases though. I'm always a bit surpriced when I meet people that tell me one should vote for the party that has the best policy for them selves personally. I've always thought of these things in terms of what will make the best society. (That doesn't mean my personal gains wouldn't affect it, but not in general or conciously)

[@jimbrookhyser]

I hear you man. I'm a little less skeptical that people voting strictly for their own best interest would be worse than what we have now. Selfishly, I want to see a cure for cancer (that I can afford), good roads, and well educated people with which to interact on a regular basis. That said, the people who voice these views in public (often as "I ALWAYS vote to lower my taxes") do tend to be intolerable jerks. Gives the idea a bad look for sure.

[@msobota4080]

Imagine that maybe the majority of people don't feel as if they receive any particular benefits from voting on long term policy in an economic and state system that is surprisingly quite dependent on people adapting in large movements in short periods of time. Maybe people just don't want to be held to some kind of impossible moral standard. "Oh aye love, remember to check the local MP's policy history before you get to your shift. It's all very significant that." Maybe the majority of people just would like to feel responsible and significant for real change in their own lives instead of asking for permission to not have all their money taken by very very intelligent people who are going to end up concluding something along the lines of "Hey guys, we just need you to spend more money. Just boost the market up a bit for us." We bloody well could if you didn't take the money, now couldn't we?

[@jimbrookhyser]

@@msobota4080 just to be clear. I get the sense you're trying to offer counterpoint to what I said. Do you identify with the description: person who always votes to lower their own taxes? Are you saying everyone should vote that way for the benefit of society? Are you trying to argue that people who say that don't come across as jerks? Maybe you were making an argument relative to the original commenter?

[@saqibmasood501]

The Proposition made at 0:54 seems to be incorrect at first as we can see that A,B and C all three are at 1,2 and 3 position at some time....but one thing that should have been clear in the video is that if a Policy that is at 3rd position it is vetoed for . Just try to think with respect to voter and not with respect to Policy

[@Stan_in_Shelton_WA]

exactly, my post - I totally disagree with your opening argument on the A B C vote. All 3 choices have identical vote results. All three have 1 vote in the #1 selection, 1 vote in the #2 selection, and 1 vote in the #3 selection.

[@Harevald]

It's ordered list of preferences, where first guy likes A more than B and B more than C. 1>2>3 in every case. So if you ask first guy what he would choose between A and B he would go for A. 1>2 If you ask second guy the same question, he would rather pick B over A. 1>3 If you ask the third person that question, he prefers A over B (lesser evil, better have something I don't prefer than something I hate). 2>3 So overall score is two voters for A and one for B.

[@bad-at-art-games]

gets even funkier, when voters also assign value to making something less desirable for a faction of other voters :D

[@frentz7]

Well done! Nice and clear presentation. Up to minute 4 so far .. I'm a math prof, have just once or twice had the pleasure of teaching this material as part of a semester.

[@frentz7]

one technical point .. the circles are (I guess?) viewed at a perspective-angle, from the sky? anyway they seem distorted (06:04)

[@dontclick7389]

@@frentz7 The camera angle is from a perspective that isn't precisely facing down. That's why the circles appear ovular. You can tell by looking at the axes, and how the width grows as the lines draw closer to the camera.

[@frentz7]

@@dontclick7389 yes, exactly .. viewed at a perspective angle, was my choice of words

[@Pumperknuckle]

This video was incredibly useful. Thank you for the learning

[@GeorgeCowsert]

It's situations like this that are why nations like America were intented to work as Republics, where the Federal government is only the mediator between foreign affairs and disputes between States. Then, each State is to be further separated into towns and cities and counties, with their relevant regulations. Policies are primarily dictated by the needs of the locals, with the policies becoming more broad and less frequent the higher up you go. Unfortunately, this concept has mostly been ditched by politicians and the general public, despite working out better for everyone involved.

[@tonyisnotdead]

and enforce on all states the constitutional rights

[@ladripper47874]

0:53 But wouldn't A and B have the same amount of votes? if you give their preferences numbers (2 for prefer, 1 for like and -1 for dislike) and add them up, each of the 3 different votes are 2:2, so none actually win

[@JessyIsMute]

yeah im not understanding this aswell

[@ravindrasinghdangi6931]

yes i also have same problem , how A is prefered over B dispite having same likes and dislike

[@matthewmoulton1]

The example does not used ranked voting. As a result, A has two votes (top person and bottom person) and B has 1 vote (middle person).

[@mactheo2574]

Confused me a bit as well. As Matthew said, top person already chosen A and cannot choose B.

[@olartio2185]

But you vote only one policy so person 1 votes A even if he likes B , person 2 votes B and person 3 votes A despite it is his second choice. Once you vote you eliminate the other options . You are right if you would assign points in a voting system

[@MrNikMielche]

Why would any of the policies win in the initial example? All of them have the excact same votes - one prefered, one ok and one against. Not one of them are bigger than the next.

[@medicineformelancholy9033]

This needs to blow up, even more than it already has. Thanks for the insight, brother.

[@prerakchoksi2379]

0:35 how A wins over B? A and B both have 2 votes

[@BS-bd4xo]

Amazing video again. Idk how far the implications in the real world go tho, cuz many factors are ignored.

[@donaldham308]

This honestly makes so much sense in so many ways.

[@nf6386]

Thanks, this is fascinating. I’m wondering if the incremental steps would necessarily be faster/fewer to reach the extreme policy if voting were not compulsory, so that within a certain narrow range of difference between alternatives, voters don’t bother to vote at all, hence cede more power to those who feel strongly/see a larger incremental benefit.

[@conlon4332]

But this worked because most of the people who voted for policies only slightly preferred them. If they don't vote at all then others' strong preferences against might win out, possibly allowing for a result closer to everyone's preferences and making it harder to manipulate. I could be wrong though, if I'm honest I am very tired right now.

[@nf6386]

@@conlon4332 it’s certainly complicated. I guess I’m thinking about national elections where there are multiple policies and values being considered, but many voters are focused on only a few, so end up effectively voting for other things unintentionally or without much consideration. In Australia, we’ve had compulsory voting for a long time and to us it seems like a no brainer to ensure that almost everyone casts a vote, on a Saturday, rather than allowing bad weather or weekday polling booth access to affect the composition of voters. We also have much more moderate, centralist politics in general (covid lockdowns aside, which have since been unwound). But I don’t know if there’s any genuine causation in that relationship.

[@ChannelMath]

I see some comments on Ranked-choice voting. I'm a supporter of that, but viewers should look up "Arrow's impossibility theorem", which shows that any voting systems have a fatal flaw (usually this non-transitivity problem). What bothers me is when people use these facts to get the public to throw up their hands and declare activism useless. Maybe you could do some more videos in that vein? I find this idea of "multidimensional voting models" pretty silly (for the reasons you say), even to the point of being just an over-complication of what we already know from studying ordinal preferences. Maybe it's useful as a visual aid.

[@1ucasvb]

Arrows's theorem only applies to ranked voting methods anyway, and it's not even that relevant in practice. The whole reductionism to ordinal preferences has been a huge disservice to social choice theory, and has produced no useful results. We should instead be studying these things under decision theory and frameworks of choice under uncertainty.

[@ChannelMath]

@@1ucasvb what is a voting system in use that is not "ranked" in the sense of Arrow not applying? I agree that reductionism to ordinal preferences is not realistic in terms of actual human preferences, but how do we do social choice then? Can you point me to some resources? (not rhetorical questions, I want to learn) thanks

[@jimbrookhyser]

I'll offer up sortition as an improvement that simplifies the process AND improves representation, rather than adding complexity but improving nothing.

[@1ucasvb]

​@@ChannelMath Approval voting, score voting, STAR Voting and Majority Judgement (but either one not fully), evaluative voting, quadratic voting are all examples of cardinal voting methods which evade Arrow's theorem. They are used on smaller scales and in various organizations, but are still not in wide use for political matters, unfortunately. Evaluative voting and approval voting have been used in large scale elections in the past, however (Venice, Greece). Approval Voting is currently in use in a few cities in the US (Fargo, St. Louis). Wikipedia has some information about usage on each of these method's articles. To understand why these methods evade Arrow's theorem, let's make something clear: the theorem isn't about voter's opinions, it's about "voting rules". It's about ballots and their tallying rule. The theorem only applies *after* ballots have been cast. The distinction is important, as it is not apparent if one assumes only ordinal information exists or is useful, and the opinions get conflated with the ballots. This has led to a lot of confusion as to why cardinal methods evade it, as people seem to think that modifying the ballots after they have been cast is functionally equivalent to running a new election with a different set of candidates. (Under a decision theory framework, it's quite obvious this is not the case, as you completely change the whole context of the election, and all the uncertainties and risks involved.) To keep things clear for others reading this, Arrow's theorem states (in one possible statement of it) that no ranked voting method can satisfy these three criteria: 1) Unanimity (if everyone prefers A>B, A wins) 2) No dictatorship (no single voter overrules everyone else's votes) 3) Independence of Irrelevant Alternatives ("IIA", the relative result between A vs. B is independent on how voter's rank a third option C). Criteria (1) and (2) are usually not an issue in practice. The "problematic" one is IIA (3) for ranked methods, as if you get everyone's ranked ballots after they have been cast, then drop one candidate, the final results between the remaining candidates can change, even though the ranking information between them was left intact. (Wikipedia's article on IIA has many examples.) This ultimately happens because ranked methods rely on majority rule, which breaks down with more than two candidates. Here's a simple example, involving a Condorcet cycle where every candidate "beats another" by a majority, in a rock-papers-scissors sort of cycle: 25%: A>B>C 40%: B>C>A 35%: C>A>B Majoritarian methods have to figure out who to pick as the winner, somehow, from this information alone. But no matter who is picked as winner, eliminating one of the other two "irrelevant" candidates flips the results the other way. For example, if B was the winner with the original ballots, the 60% who prefer A over B would elect A if C were dropped from the ballots. So the results flip from B to A by dropping C here! Basically, no matter how you try to replace the concept of "majority winner" here, it can contradict a reduced 2-candidate scenario's "obvious majority" picture. Now imagine instead you have a cardinal method. The same ordinal information above could arise from many possible underlying opinions (which MUST be cardinal, that is, they must involve some "strength of preference", as humans don't think in ordinal terms and denying that fact is ludicrous). As an example, let's say these cardinal opinions were ultimately mapped, somehow, to the following score ballots (I'll use Score Voting just for simplicity!): 25%: A=10 B=8 C=0 40%: A=0 B=10 C=3 35%: A=5 B=0 C=10 The final (average) scores would be A=4.25, B=6.0, C=4.7. B wins here too. Now if you drop C from the ballots and compute the average again, nothing changes: A=4.25 B=6.0. So C is completely irrelevant for A and B *within the ballots*. This will be true for any scenario or candidate dropped, which means score voting obeys IIA, as well as unanimity and non-dictatorship. Thus, it violates Arrow's Theorem. But as I see it, coming from a decision theory picture of voting as opposed to ordinal social choice, IIA is a meaningless criterion. Modifying the ballots after they have been cast is not really that useful or important in practice, and since voters cast ballots (ranked or rated!) based on risk assessments, based on the comparative evaluations of every option available, and not through "absolute opinions about candidates, existing in a vacuum", this means voters opinions themselves necessarily must violate IIA, if they are to be meaningful in an election, or even in aggregate. That's why if you drop a candidate and re-do the election, no method, cardinal or ordinal, will always "obey IIA". It's not even something we should strive for! If you want more information about cardinal methods, the validity of using cardinal information instead of just ordinal information, and why we can use them to make collective decisions, look for a top-level comment I posted in this video as I already wrote some stuff elsewhere. Many common points/objections have already been covered. If you want to learn a bit more about the underlying decision-theory approach, there's no unified resource on this particularly applied to voting (yet), as the field is still stuck on ordinalism, but I recommend: "Voting and the Cardinal Aggregation of Judgments" - C. Hillinger (Hillinger's other papers on the subject are also good) "A Cognitive Basis for Cardinal Utility" - T. Kornienko "Group Decision Making Using Cardinal Social Welfare Functions" - Keeney and Kirkwood "Multiattribute Preference Analysis with Performance Targets" - Bordley and Kirkwood "Stochastic Dominance for Decision Problems with Multiple Attributes and or Multiple Decision-Makers" - Nakayama et al. "The relativity of utility - evidence from panel data" - van de Stadt et al. "Relative Measurement and Its Generalization in Decision Making" - Saaty "Semiorders and a Theory of Utility Discrimination" - R. Duncan Luce "Cardinal Utility from Intensity Comparisons" - Shapley "Decision analysis using targets instead of utility functions" - Bordley et al. "Exceeding Expectations: Stochastic Dominance as a General Decision Theory" - Tarsney Most of these detail the concept of "cardinal utility" as a legitimate, reproducible or robust theoretical or empirical model for decision making under uncertainty, involving multiple agents and multiple attributes for the options. But this is only possible, in practice, if the induced scale is shared among individuals and pairwise comparative, which it is in the case of voting, as a matter of necessity. However, the concept of "cardinal utility" is too simplistic and carries a lot of unnecessary baggage, which gets a lot of distracting criticisms (even in some of these papers). But it's also completely unnecessary and obsolete. You can frame everything here in terms of probabilities, "targets", and the idea of stochastic dominance over subjective probabilities. Under such an approach, cardinal and ordinal voting methods become identical, as the ballot simply acts as a way to encode "probability that a voter preferes A>B given their current subjective probabilities", for every pair of options {A,B}. The goal of a voting method is then simply to "maximize the expected number of positively satisfied voters", which is hard to object to, and both cardinal and ordinal methods can satisfy, under different assumptions. But that's enough room for a whole lot of further discussion in the future.

[@drdca8263]

@@1ucasvb Gibbard's 1978 theorem still applies in general though. (Jim Brookhyser's comment suggesting sortition gets around this theorem, at the cost of the randomness and the "well what if someone votes for something stupid and by chance they happen to be the one whose choice was chosen?" aspect. It's a very elegant solution, but I really don't see it being implemented on a large scale anytime soon...)

[@anonymousn5292]

Amazing video, really interesting point and beautifully and articulately expressed with the visuals. Subscribed

[@kiryonnakira7566]

wrongs from the start. why would Policy A > B ? all other 2 statements are wrong. Let's count A : 1 prefered choice, 1 2nd choice, 1 no B : 1 prefered choice, 1 2nd choice, 1 no C : 1 prefered choice, 1 2nd choice, 1 no All are equal.

[@TunaBear64]

You don't understand First past the post it seems, which is reasonable, that system is stupid but used widely.

[@LeeSpork]

But if you are only counting two of the policies against each other, then one of the "second choices" becomes a first choice for that count (as their actual first choice isn't an option)

[@bantix9902]

The non-transitivity of majority decisions is an interesting point to bring up, although I don't see a better alternative as of now. This model you propose is obviously very limited and makes a lot of assumptions. In reality, the voters position is dynamic and influenced by a huge amount of inputs, including the current policies. As you pointed out, compromises are made constantly, therefore I would assume that more often than not, a minority of voters get policy A enacted in exchange for another policy B later down the line.

[@SmogHouseTradingCo]

Ranked ballots help this. Either way, whoever can set up what the vote is is ultimately in control of the outcome. That has been the fundamental problem of democracy. Socrates hated it because of populism and the control of what gets voted on. It's a human system so will have a weakness, but we can make it more difficult to corrupt than it is with first past the post as is.

[@bantix9902]

@@SmogHouseTradingCo I think what you describe is a problem in every political system imaginable. It's an artifact of size and complexity. Democracies often times see themselves as the best political system. They restrict the meaning of absolute free will to some subset that only operates within the narrow boundaries of hegemony, while the idea of absolute free will is like a ghost haunting every supermarket, house and mind. The laws reference a freedom but the boundaries are set up by economic realities and only indirectly through laws controlling the material conditions. It's a good ploy and ties back into the dogma of individualism under capitalism.

[@guteksan]

Some time ago I came up with a similar thought: Consider a democratic society wants to impose some limitations on the rights to vote based on intelligence. One policy is to give the right to vote to the smartest people only (say, IQ>=130, SD=15). Provided that voters are rational and vote for their own interest, such policy would lose, because most voters (almost 98%) have IQ < 130, so passing such policy would be against their interest. But consider now, that we start with a policy to cut off voters with, say IQ < 80. Such policy is likely to pass, because most voters have IQ>=80 and it is in their own interest, because cutting off some others voters reinforces their own votes in any future voting. But now in the same way we can cut off voters with IQ=110, so increasing the cutoff IQ by 10 might technically still work. But the next step can't be 120, because there is only 36% voters with IQ>=120 in a group of voters with IQ>=110. What we need to do is to simply take smaller steps, like 5, then maybe 2. In the end we should eventually reach the cutoff IQ=130, so the policy initially rejected by 98% of the society would be now accepted. Well, of course the smarter the voters become, the more aware they should be where this agenda is going to, and that they themselves will be cut off eventually. On the other hand, with smart campaign full of manipulations and blaming minorities ("this stupid mass can't understand us, the elite!") I can totally see people falling into this trap.

[@atoucangirl]

there's a wrinkle: it assumes that the policy that pleases the most people is actually the one that hurts the least people. per example, if a certain voter wants a policy whose purpose is to hurt other voters in some way (usually for their own benefit), then the "compromise policy" we started with probably isn't the best one. other than that great video, very insightful

[@crazydragy4233]

This bugged me too. Most policies are complex and always about trade of interests/resources. And not everyone actually wants compromises, some positions are inherently anti fellow voter. It's a cool basic model but it doesn't account for reality

[@dontmisunderstand6041]

No, no assumption like that was ever made. The one and only assumption being made is that people vote according to their own preferences. That's it. That's all of the assumptions.

[@jonathandpg6115]

No this is a basic extreme example of how you can get worst outcomes for everyone by having intermediate policies. That assumption isn't made it just happens to be the case because it's an easy yes/no with a single measurable benefit/con with his small example and his success case. In reality in is more nuanced but the idea is to understand the how

[@atoucangirl]

@@jonathandpg6115 "By playing voters against each other, we, the agenda setters, have caused them to agree on a policy that is much worse for everyone than the policy they began with." I'm not saying that the policy that's very far away from everyone's interests is good, but there are a lot of cases where a "compromise" policy is bad. example: racists wants black people to not have rights, other people want them to have rights. a compromise would be to give them some, but not all rights, which is still bad. I won't bother arguing further btw I made my point and I don't have anything else to say

[@yash1152]

1:13 wait, Policy A is 1's 1st, and 3rd's 2nd. so, both should be equal (if doesnt like and 3rd are equal)

[@aouerfelli]

What do you mean? Both the second voter and the third voter will prefer C to A. Only the first voter will vote for A. So C will be chosen. Because they will not ask voters how much they prefer one policy to another, they just ask them what they prefer the most. And that is the problem of such voting systems, the only thing which matters is how many people prefer some option over the others; not how much they prefer it so even if the majority prefers it slightly it will be chosen; the amount of dislike towards that option is disregarded.

[@yash1152]

> _"Because they will not ask voters how much they prefer one policy to another, they just ask them what they prefer the most."_ @@aouerfelliuhm, yeah, it makes sense.

[@yash1152]

> _"prefer C to A"_ @@aouerfelli ohw, that's the thing i missed. The comparison is between the policies of interest only, and not on the numeric order of preference.

[@thoughtsfocused2262]

The premise of the three voters voting for A B and C seems contextually inaccurate. This premise is the rock, paper, scissors paradox in disguise making this a circular argument. To say that C is preferred to A making C the second most desirable preference is invalidated by A being preferred over B and B being preferred over C and so on.

[@saqibmasood501]

Exactly that's what I thought

[@cheetosnour.scratch-learn]

I love you, spanning tree

[@CosmicKnight1]

Loved the video, but the audio was too sharp. I found myself having to turn down the volume to more comfortably listen. Great video, it’s a trope in business there’s power in setting the agenda and this clearly illustrates that, love it.

[@dananskidolf]

Voters are very bad at noticing all the dimensions or how much they're affected by them and thus cannot accurately calculate the distance of policies from them - the policy space is non-Euclidian, with a very blurry, peculiar and perspective-dependent distance function.

[@1ucasvb]

Yes. The better way to model this mathematically is with multi-attribute decision models, but they don't produce nice intuitive images. But the things in the video still apply. The Euclidean space assumption isn't really the issue, but the ranked preferences assumption, i.e., the distortion here arises from ballot space, not opinion space.

[@AnyVideo999]

Your right that a Riemannian manifold would be more appropriate where things can get far more interesting.

[@crazydragy4233]

​@@1ucasvb Also maybe I'm wrong but I feel like it doesn't account for anti voting? Where people don't vote for the best option for themselves but the worst option for others

[@theultimatereductionist7592]

This method can be used by any player against the "red policy voter", too, by picking any other issue. The issue can even be very abstract, such as "Do you support the idea that if those fight for a cause deserve to be rewarded for their sacrifice and effort?" A majority of voters will say they do. Now worker unions are free to fight for what they want and be rewarded. So too climate activists. So too prisoner rights lobbyists. So too animal rights vegans fighting the holocaust of factory farming. So too Antinatalists. So too UFO disclosure lobbyists.

[@l.o.x.i.1768]

good video. solid reasoning yet fairly comprehensible

[@alexyoung6418]

Great video. I always had a hunch that there should be a way to analyze these things mathematically but I suck at math. Now I see what exactly one of the methods is. Always a good thing to discover one more hidden possible way of making mistakes when common sense tells you otherwise.

[@bojackhorseman4176]

It's fascinating, isn't it? Twenty years ago, if you asked most people, they'd say that art has an inherent human aspect that can't easily be replicated by machines - or at least some variation of that. Yet now, look at what AI paintings and text can do, some might call it soulless but it's definitively better than any scribbles I can draw. To think that such intrinsic aspects of the human condition, like our imagination and behaviors, can be quantified and manipulated by numbers... I wonder what else we can fiddle with next?

[@alexyoung6418]

@@bojackhorseman4176 Indeed. A lot of the things we didn't think was possible recently turned out to be limitations of computational power. Even with stuff that are mathematically proven to have no analytic solutions or stuff that would have taken more input variables than available in field to solve, there are still engineering alternatives to tackle them with blunt force computations or less-than-perfect trials and errors. Fortunately most of the practical applications only need to deal with a finite domain. Hope these potentials offer a good opportunity for us as humans to reflect upon ourselves and start identifying our true core values and false basis of our self-importance. A lot of people couldn't conceive a world where we humans were not so special, as though without a unique touch from God, we would have been rendered unimportant in this world, even to ourselves. Well, maybe it's time to toughen up and accept that we may be neither too special nor important to the world after all, but we certainly still are important to ourselves, not for any glorified reason, just because we are ourselves, and that we want to become better at it, that alone should suffice.

[@pete5516]

@@bojackhorseman4176 the reason it’s ‘better than some scribbles’ is because AI doesn’t actually learn like humans do. An AI constructs a piece of art essentially by tracing other art it has been trained on, and modifying it slightly until it can be called ‘new’ or combining it with enough snippets of other pictures that it is unrecognisable. This is different from how humans learn because we don’t need to directly trace things, we can call upon our collective experience and have a general idea of what an object should look like. While it’s called ‘AI’ what most image generating AI do can barely be called intelligence, and it certainly isn’t machine learning. The real AI is used for military intelligence or other things far more important and lucrative than art. Most ‘AI’ art programs are posing as AI while actually not being that at all, but they can call it AI because there is no real definition yet for what that means from a marketing perspective, and it’s also very hard to prove whether or not it was actually made by AI.

[@wubbledub]

Of course the problem still remains of people not understanding the long term ramifications of policies. A how a policy may ultimately be in their best interest but they have been led to believe that it is harmful to them.

[@Banana_Split_Cream_Buns]

- Ranked Choice Voting. - Mixed Member Proportional Representation for Lower Houses - State or Regional based Representation for Upper Houses, that may or may not to some degree equalize the representation given to each state, province or region (depending on the nation's size, physical geography, and historical cultural disparity) - Collegial _de jure_ Executives (whether this is the _de facto_ Executive too or not) that are elected from a joint sitting of both Houses like Switzerland, or popular vote, yielding a proportionally representative multi party makeup - Appropriately devolved (i.e. autonomous to some degree regional government). - Citizens' Initiated Referenda (at a significant threshold of voters' petitioning). - Census questions to include non-binding survey questions on policy related matters with 0-9 scalable answers. - National elections held every 3 years. - A Constitutional Bill of Rights for fundamental human rights. - A Legislative Charter of Rights which can expand on the Bill of Rights for more detailed matters. - A sovereign currency and sovereign nation.

[@dmsdozent5273]

Isn't there an error at 1:18, where you say policy C would win over policy A and high light booth policy C votes. If you look at the policy A votes they are the same. So actually none would win or am i missing something?

[@thecarman3693]

Nope ... not missing a thing. That's exactly what I was about to post before looking to see if someone else noticed the obvious. C is not greater than A, they're equal. Thanks!

[@laconicdraconic697]

They are the same. However the policy C takes precedence over the policy A for person number 3 since it goes C A B which means despite the fact policy A is the second prefered, the choice the person chooses policy C when give a choice between A and C. It is an either or so one must be picked over the other. Person one chooses policy A. And person 2 chooses policy C since their primary choice policy B isn't available.

[@AaronMichaelLong]

This example is *precisely* why we're a Republic, not a Democracy. Laypeople do not understand the law. Their perferences are the result of emotions, not informed study of an issue. The point of an election is that you're supposed to choose a representative whose judgment you trust, so that they have the education, expertise, and time to study and issue and pass an intelligent policy. Unfortunately, the Founding Fathers did not anticipate mass communication, and the rampant demagoguery it produces. Back when the Constitution was signed, there were approximately 15,000 citiizens per Congressman. Now that figure is closer to 700,000. The point being, it used to be possible for you to know your representative personally. He was a figure of import in your community. Now all we have is 435 geriatric hairdos with staffers running their Twitter feed.

[@FaliusAren]

Do you have any examples of this issue somehow becoming exploited in reality? Given the large number and huge variety of intermediate policies in your example of the phenomenon, it seems vanishingly unlikely anything like this could actually occur.

[@jaideepshekhar4621]

Hitler's rise to power?

[@antoniusnies-komponistpian2172]

​@@jaideepshekhar4621 Wow, this actually makes sense.

[@aduckwithgrapes9572]

@@jaideepshekhar4621 thats actually an amazing example lol

[@reuternopalzin2422]

Arrow's theorem (which the first part of the video belongs to) is always used as "democracy cannot exist" kind of what this video did (except that the conclusion is something pretty cool, look for consensus). However this kind of scenarios are EXTREMELY rare for big numbers and they don't appear if you look at all the election held throughout the last centuries around all the globe, and this is something that politic and social scientists always point to. Also that only happens in ranked elections with more than 2 options. So, the scenario at the beginning of the video is something that you are not going to find in elections with big population where 1 person-1vote. (Even if it is a ranked election)

[@dontmisunderstand6041]

@@reuternopalzin2422 You must realize that your assertion at the end there is definitively disproven by a grade school level understanding of 20th century history, right? Or, even without that, having functioning sensory organs at any point in the past 10 years.

[@fejfo6559]

My idea to fix the voting system so this becomes impossible is to let voters "save" votes (eg everyone gets 10 votes for the next 10 policy changes and they can choose how many votes they spend each time). This way the voters that gain little would vote a little bit in favor and voters that lose a lot will vote strongly in favor.

[@dontmisunderstand6041]

One problem there is that this rigs elections in favor of demographics that already benefit from the status quo. Rich old people already rule the world, this would just make it harder to change that.

[@Redpoppy80]

I don't think a democracy can ever avoid the problem of self-destruction because humanity itself is the issue in the longevity of democracies as a concept and we have too many adverse behaviors. Pride, Greed, Jealousy, Wrath, Prejiduces, weaponized ignorance and worst then the issues is that any solution can simply be adapted around in favor of the anti-democrats and ultimately they will always win. History is very clear, Democracy will ALWAYS fail and fall to authoritarian rule. Frankly, we need a system that goes beyond democracy if we want to preserve humanity.

[@dontmisunderstand6041]

@@Redpoppy80 One thing you may not have considered is that true democracy hasn't really been tried on a scale beyond the size of a small town. The issues with this system are associated with negative aspects of a republic, not necessarily all democracies. History has no opinions on democracies, because they've never existed in any meaningful capacity. It has, however, shown consistent proof that republics inevitably become imperial dictatorships. Quite literally at a 100% rate.

[@brentt6714]

Good job highlighting the problem of the US' culture of hyperindividualism without triggering a defensive response.

[@phos5990]

the area of "overlap" gets bigger over time, and further away from everyone. very interesting. the take away for me is we need to care about others whose interest isn't necessarily the same as our own group. inclusivity helps everyone.very well presented video.

[@bimbumbamdolievori]

doesn't feel right to me. In your initial explaination you consider the *first* 2 votes of 2 ppl, and the *second* vote of the 3rd one in order to determine if A > B etc.. but this is a biased perspective (and question). You must ask which across A, B and C got more votes (this is the correct question! applying ">" and "

[@calliopehu1924]

huh? no, the question they're asking is "if these three voters each got a ballot with the two given policy options, which policy would win?" On this front, each voter would individually have one vote, and that vote would go towards the policy that they most preferred of the two given. the listed ranks of policies for each individual voter helps determine which policy that voter will vote for when given a choice of two. and the greater/less than is just to show, between two policies, which got more or less votes, which I think is an entirely fair use.

[@alan5506]

So the problem isn't qith democracy but with first past the post. I already knew that. Video is useless and misleading.

[@Mikhanius]

@@alan5506 it's not a problem with just FPTP, a similar issue appears in many ranked choice voting systems. You could also look into a related problem with the Condorcet method, which considers all the possible 'agendas' (pairs of candidates) and chooses the one that wins in each pair, but it is not always the case that such a candidate exists, sometimes a rock-paper-scissors situation like shown here may arise, even if it is rare.

[@calliopehu1924]

@@alan5506 No, the same results would have occurred in a system with Single Transferable Vote (which is the same as First Past the Post when you only have two choices available). The best solution would be to have all three options on the ballot at once, which would have resulted in a tie in both FPTP and STV. But that just proves the video's point, that power lies in agenda-setting just as much if not more than in voting.

[@alan5506]

@@Mikhanius Ah. That's a very interesting then. But another issue I have with this is, is it even possible to use this? The world is chaotic with many unknowns. Can politicians really model the population so well as to use this strategy? I have no doubt the answer in the near future will be a resounding yes, but has it already happened?

[@ResonantTonalityMusic]

Oh blimey, I barely remember preference schedules from discrete math but I recognized it immediately and felt a deep concern.

[@Nick-ij5nt]

This is pretty interesting in theory but I think this actually neglects how the real world operates, or at least how it currently operates, maybe the world used to work more like this. There generally isn't a list of policy options to choose from. For most major policies there's usually only two options that are diametrically opposed to one another, with little to no compromise from either side.

[@crazydragy4233]

I feel like that's a very American take. Many places have more than just 2 major parties My issue with this is that policy is too abstracted in the example. If we understand it as moderating resources and power, and policy changing as who gets what and when, then we got to account for moments where policies are about disempowering and empowering specific groups at the cost of others, and not over all interests of everyone. Some policies are by design anti cooperation and going against the interests of others/voting without them.

[@Nick-ij5nt]

@@crazydragy4233 Yes, I was referring specifically to America. I know other countries have parliaments with multiple parties in them. So yes, this video is probably more relevant to Europeans than Americans.

[@dontmisunderstand6041]

... by definition any system that doesn't work that way cannot be called a democracy. When you cast a ballot, you are very much doing exactly the thing this video's model is describing. Whether it's a policy or a candidate is irrelevant. The number of available options is also irrelevant. The number of preferences being taken into account is also irrelevant. The literal only part of democracy that is relevant to the video is the fact that people vote in their own self-interest. Also it's not a theory. It's explaining to you how basic math works. There's no theory involved, it simply is correct by necessity.

[@elpsykoongro5379]

USA isn't even a democracy, it's more like the Roman republic

[@IowaKim]

This is the video that policy makers don't want the average person to understand. Thanks for explaining.

[@michaelbyron1166]

One of the best , most informative videos that I have seen in years. Even thr commentary here is very educational. Thank you to everyone.

[@greenjacket4605]

This kind of logic supports an idea I've had which I call the 'false dialectic' where people's political opinions are influenced by the issues that are presented to them and this forms the basis of the red-team/blue-team trap a lot of people in American politics find themselves in.

[@jonathansmith3889]

Tbh that’s more common sense. You’re less likely to donate to a foreign charity than a local one. You’re more likely to help a friend than a stranger. Why? Because it’s basic psychology my boy, but what you’re just stating is more of a fact.

[@MrMrtvozornik]

@@jonathansmith3889 Yeah, it's called reciprocative altruism and it's how societies are built, which is also why it's very hard to keep them coherent once they are big enough. Which is why democracies never have the largest powers nor land mass, compared to empires and other forms of governments. But then again, let's not confuse democracy and demagogy, and ancient Greeks who made democracy would call modern democracy a demagogy. The idea that somehow a mentally unstable foreigner(first generation immigrant) should hold same voting power as a native university professor is preposterous and contrary to human nature. The same contradiction is seen in ideas such as getting some random country in Africa help to stand on their own feet while you have massive homeless problem in your own country. Asking people to care about other people on the opposite side of the world while they see suffering in their own neighborhood is beyond idiotic. The same way one ought to fix themselves before they try to help others, countries should take care of their own before they start massive humanistic endeavors across the globe. Aaaaaaaand I just got called a N*zi for this stance. Slippery slope for the democracy stans. People who say that seeing world in ''us vs them'' mentality is somehow wrong are delusional, given enough time one could actually divide the whole world in sub categories of us vs them. The issue is, one shouldn't just blindly stand on the side of ''us'' while blindly hating the side of ''them'', but lets be real, majority of people won't follow that logic, despite it making sense. tl;dr Only way to see whole Earth unified is to offer another ''them'' to rally against, be it Mars colony in the future resulting in Earth vs Mars stance, or aliens or w/e, but humans are divisive by nature. Which follows biology because guess what, when push comes to shove, your ability to procreate is unlimited, while resources are limited on this chunk of rock, so there's always going to be some friction over who's offspring is going to get the resources needed to survival/leisure etc.

[@giannicettolo308]

Amazing video!

[@CobraTheSpacePirate]

That pretty much aligns with the "Nash equilibrium", the governing dynamics model that was talked about in the movie, "A Beautiful Mind".

[@xcheesyxbaconx]

It's an interesting idea, doesn't really apply entirely in the real world though because the average person prefers the status quo over something that would only marginally benefit them, due to human nature. Or at the very least they wouldn't bother to turn out to vote. While those minority that are heavily against the new policy would be more likely to vote. It's definitely a good argument against mandatory voting though.

[@mortophobegaming6454]

"Democracy is the worst possible governing system. With the exception for all the others" - Winston Churchill

[@賴志偉-d7h]

Clever but meaningless smart talk.

[@지성준-m7h]

astonishing and mind awakening criticism...!! Thanks alot. It gives me a much wider vision to the world.

[@firstnamelastname9237]

Fascinating! I do want to know about actually involving the order of preference in voting. So, if your favorite got 10 points, second 6 and least, none, how would that change the results?

[@dontmisunderstand6041]

It wouldn't change the results, only make it harder for an average person to understand the concept being displayed here. Each preference is represented by an axis, and in reality there would be an infinite number of axes. Any finite number of policy choices would fail in exactly the same manner as a binary choice system. The lesson here isn't that we need to be presented more options in our pursuit of self-interest. The lesson is that acting in self-interest is guaranteed to be self-destructive in the long term.

[@underrated1524]

​@@dontmisunderstand6041 Voting for the benefit of everyone doesn't do anything to solve the problem. Different people have different ideas of what "benefitting everyone" looks like, and the agenda-setter can use that difference to implement the strategy indicated in the video.

[@dontmisunderstand6041]

@@underrated1524 "Self interest is bad" doesn't automatically mean "selflessness is good". Though, there's a flaw in your reasoning as well... the idea "different people have different ideas about what's best, therefore voting for what they think is best won't solve the problem" isn't actually describing a scenario where people vote selflessly. It's still voting in self-interest.

[@underrated1524]

@@dontmisunderstand6041 If we could count on people to choose what is ACTUALLY best for everyone, rather than what they *think* is best for everyone, we'd barely need a government at all.

[@dontmisunderstand6041]

@@underrated1524 Precisely.

[@SuicidalSummerSnowWoman]

seems pretty obvious when you explain it...good teacher